Askey's Sony F828 review

[Cadenza]

FYI:

It was about time, but Phil Askey's review of
has at long last finally been published at dpreview
as of this morning. FYI, the link:

http://www.dpreview.com/reviews/sonydscf828/

It seems like Phil is working hard on validating
the rumors that he is on Canon's payroll... Else,
the F828 is just a lousy camera...! :D

Regards to all, Cadenza

[msnow]

This is a much more thorough, empirical and realistic review than the one I read by Michael Reichmann at luminous-landscape.com a couple of weeks ago. I think he did a good job on this.

[Anders Östberg]

To me it seems a shame this camera's user interface isn't matched by the image quality.

I'm sure it can be used to produce great photos just the same, you just need to do more post-processing work than I would expect of a new digital camera.

Maybe next time they'll have a winner.

-Anders

[Pekka]

And what does this thread has to do with Canon EOS? :?:

[Cadenza]

Pekka,

Quite a lot actually. The Sony F828 is being marketed
as a direct competitor to the D-300 and the G5; trying
to reach the same target consumer population. A lot of
us I trying to decide whether to upgrade into 300Ds or
the Sony F828.

Otherwise, thanks for keeping this fabulous site.

Cadenza

[Anders Östberg]

And what does this thread has to do with Canon EOS? :?:

Probably meant to make you feel good about using an EOS... ;)

[Pekka]

Pekka,

Quite a lot actually. The Sony F828 is being marketed
as a direct competitor to the D-300 and the G5; trying
to reach the same target consumer population. A lot of
us I trying to decide whether to upgrade into 300Ds or
the Sony F828.

OK.

[Guillermo Freige]

Let's play devil advocate here :)
I've a DRebel and an S50 (who utilizes a Sony 5MP sensor), and regarding noise issues, I think Phil review is telling only half the story.
The Drebel CMOS sensor has a very low noise in midtone to highlight areas, even at ISO 1600, but at the shadows things start to get noisy even at ISO 400. Sony CCD sensors have a more lineal response. Of course they are noisier at the shadows, but the noise bulidup in highlights is much more noticeable than in the CMOS case. Also as noise increases at darker values, if you don't test images with the same brightness you are favoring the lighter one.
I noticed that in my own cameras, and I've used ISO 800-1600 a lot in the DRebel, so I've experienced this different noise characteristics in my own work.
As Phil is using a light grey patch to the noise test (the standard practice), it's using the CMOS "sweet spot" regarding noise. Also, as the F828 generate darker images, also the noise level is increased. I'll be nice to see the same test with both cameras exposing the patch at same level and using a dark grey patch. I think the results, still much favorable to the dRebel, will be not so dramatic as the "dRebel ISO 800 is similar to F828 ISO 100 noise levels" situation.
Also, regarding Michael Reichmann conclusions, I think the guy is right. "Scientific" tests are like estadistics. Very useful, but to some extent. They aren't the revelated true, but just numbers. And after all, the only thing that matters is the image, right?. How well it looks in the final output medium (screen or paper). So If it looks fine there, it's a good camera. After all film isn't noise free (we use to call this noise "grain", remember?) and probably at the same level than F828 noise, and there is a lot of people happy with the film "noise" :)

Just my 2 cents.